Christine Mummery, Leiden University Medical Center

Some patients with heart failure have been treated with bone marrow and other adult-derived cells. Benefits so far seem modest and temporary, and the mechanisms are poorly understood. Human embryonic stem cells can become cardiomyocytes and could perhaps be used to repair heart tissue. Christine Mummery, who is moving to Leiden University Medical Center, in Leiden, the Netherlands, describes efforts to use cell therapy to heal failing hearts1.

(Note: This is part of a series of interviews conducted to accompany Nature Insight Regenerative Medicine .)

How would cell therapy for heart patients work?

One question is how can we replace cardiomyocytes, which is the major cell type that gets lost. Can you make cardiomyocytes from bone marrow? Turns out you can't; nevertheless, bone marrow seems to have some very positive early effects.

It could improve vasculogenesis and prevent cardiac damage; we might be able to find ways to get bone marrow cells to do these things better by selecting the right cell subpopulation, things like that.

What's very clear is that it doesn't seem to be dangerous. In Germany, you can spend €5,000 US$7,700 and get your own bone marrow in your heart, but as far as I know there is nothing that gives long-term benefit to the heart either in mice or in humans.

You can put cells in and say "Will they help the heart?" and a lot of cells will do this, at least temporarily. But we don't really know how these things work.

What about making cardiomyocytes to replace the dead cells?

I think the journals should set down guidelines about what is a cardiomyocyte. A lot of referees will accept anything that can improve the heart.

At the moment, I would say there are two types of human cells that can form cardiomyocytes: human ES [embryonic stem] cells and these cardiac progenitor cells, which can be derived from biopsies. Because these cells are cardiac progenitors, you get cardiac cells, you get endothelial cells, you actually get human vessels within the human graft in the mouse; the mice do better but still get heart failure.

Almost any cell will cause functional improvement in the heart of a mouse with a myocardial infarction. The effects might last two weeks, three weeks, even a month. That doesn't say anything at all about whether they are heart cells or not., Christine Mummery, Leiden University Medical Center

Almost any cell will cause functional improvement in the heart of a mouse with a myocardial infarction. The effects might last two weeks, three weeks, even a month. That doesn't say anything at all about whether they are heart cells or not. I've never been convinced by any data showing that mesenchymal stem cells can make bona fide heart cells. What happens is that they do show some markers of cardiomyocytes, but they don't show sarcomeres [protein bundles characteristic of muscle cells] in convincing amounts.

When you get cardiomyocytes, what happens in the heart?

There are issues of integration in the heart. We see the human cardiomyocytes lining up in the mouse heart, but they are isolated from the rest of the heart tissue. If they really coupled, the human cells probably kill themselves by tachycardia; mouse cells beat 500 times a minute. They have a protein-matrix layer separating them, which means you've created an obstacle to conduction. It's quite risky; we're putting electrically active cells in, but they are also immature, so they can behave like pacemakers. We have to be sure that we don't send the whole heart haywire.

What are some of the other hurdles to introducing cells for therapy?

Oh, many. How do you get enough cells in? The heart is beating, so it squeezes cells out. We lose a lot of cells that way.

How do you get them to align properly? When you look at the heart, cells are aligned in sheets which are placed at certain angles with one another It's an absolutely beautiful structure, and you have to recapitulate that.

We still need to find the right delivery method. Should we inject deeply? Should we inject shallowly? With a liquid matrix? With a sac or gel?

What's the timing? We don't even know that for bone marrow [injection]. Do we do it quickly or after a week when the inflammatory response has started going down, or after a month when the actual recovery process is happening?

The timing issue is something that is very difficult to address in many countries because the regulations do not allow you to perform surgery on the same animal more than once. So you've got limited opportunities to do these experiments and make observations.

How can animal studies be improved?

That's where you find the disparity between basic scientists and cardiologists. We do these very clean wounds in the hearts of mice; a normal heart patient would be stented very quickly after, so the vessels would be open. We tie the vessel off permanently and then we put in cells immediately.

In an ideal world, you could take a large animal like a pig or a sheep, you could pinch off a vessel temporarily and then release it, put in a stent and then put in cardiomyocytes.

Those experiments are not being done, largely because it's very expensive. A pig can cost ten to twenty thousand dollars; that's compared to, say, $50 for a mouse. We'd never be able to do these experiments in primates in Europe.

Why are you moving to Leiden University Medical Center?

I thought it would be more appropriate to have our work in a research hospital so I will have access to clinicians. The types of questions that basic biologists want to answer are sometimes not exactly what clinicians want people to do, and it's very good when you can find out that what you might research might have little effect in the clinic or that what the clinicians really want to know is X, Y or Z.

The old heart of a patient who receives a transplant is really so terrible. It looks all bruised with bits of hard tissue. The vessels are so brittle with atherosclerosis [that] you could almost snap them. I imagine that little bit of scar tissue in a mouse that I might be able to cure one day and think 'I'm so naive'.

It makes you realize that heart transplantation is still the closest to a cure for this disease. We're really looking for specific subtypes of the disease to treat.

If you were writing a review on transplanting cells into the heart in ten years' time, what do you think it would say?

I think we will have found ways to stabilize vasculogenesis in the earlier stages. We will have found ways to best prevent damage from occurring. We will have figured out ways to make cells make heart tissue, and we will be close to using them to actually help repair the heart. The challenges would still be long-term immune rejection and perhaps dealing with heart failure that had already started.

What will you do with embryonic stem cells in the meantime?

Many of us think that cells will be great for therapies in 20 years time. But what ES [embryonic stem] cells are going to be good for in the short term is testing for drug therapy. What they use now is dog myocytes in culture, and those work well but they cost you a lot of dogs. And there are serious ethical issues there. Human cardiomyocytes would hopefully have a better prediction